We have developed a semantic registry that can be used to discover sensors, sensor networks, services and data sources given an appropriate description. This implementation relies on an extension of RDF (called stRDF) that allows describing spatio-temporal metadata for every resource that can be potentially used, and a query language that extends the W3C SPARQL recommendation with spatio-temporal capabilities (called stSPARQL).
The underlying implementation of this registry is Strabon, a storage and query evaluation module for stRDF/stSPARQL. Strabon is built on top of Sesame and extends Sesame’s components to be able to manage thematic, spatial and temporal metadata. Some of these extensions are:
- The Sesame’s Storage Manager is extended to be able to store spatial literals in PostGIS, which are expressed as linear constraints (in the style of constraint databases) or in the OGC's Well Known Text specification (used by most spatial DBMS). This extension also implements libraries that deal with geometries and the conversions that take place during data loading.
- The Sesame’s Parser and Evaluator are extended to handle stSPARQL queries. Currently, in addition to standard SPARQL operations, Strabon can perform spatial selections, spatial joins, spatial projections and construct new spatial objects from existing ones.
- The results of stSPARQL queries are formatted according to the SPARQL Query Results XML Format, where spatial data are expressed using the constraint representation or OGC’s Well Known Text representation. Results can also be encoded in Keyhole Markup Language (KML) which is an OGC standard so that they can be easily presented in a mashup.
For further information, please contact